a a 
oo 
=, 
1906] OLIVE—DEVELOPMENT OF EMPUSA 199 
filaments, which in a young condition may have but few septa, while 
in a later vegetative stage they develop numerous septa. As the 
larvae crawl about over the surface of the slide, the hyphae can be 
seen sometimes extending the full length of the body-cavity, some- 
times copiously developed in the posterior portion only, again more 
abundant in the anterior region. As the body parts move, or the 
blood flows in the cavity, the mycelium shows. corresponding pul- 
sations and movements. Fig. z showsa ‘portion of a branched 
vegetative hypha, taken from such a larva which was crawling just 
below the surface of the dung culture. Evidently the vegetative activ- 
ities have here almost ceased, and the 2-, 3-, or 4-nucleate cells are 
about ready to send out the radial conidiophores. Fig. 2 shows a 
section of a younger hypha, in which the cells contain a varying 
number of nuclei; while fig. 3 represents a still younger stage, the 
- earliest condition within the body of the host which I have succeeded 
in obtaining, growing in a larva in which the mycelium was composed 
of but a few scattered filaments. The section of the hypha, of which 
jig. 3 represents only a portion, shows in the preparation twenty-two 
nuclei, but in the entire length, as far as traceable, not a single cross 
partition. 
While in my own investigations of the vegetative stages of Empusa 
sciarae there remains as yet a small gap, from the penetration of the 
infecting hypha to the production of such multinucleate mycelial fila- 
ments as are shown in fig. 3, I feel reasonably certain as to the method 
of procedure. Germinating conidia, growing in a sterilized decoc- 
tion of cooked larvae, are shown in figs. 4-9. Figs. 4-7 illustrate 
the germination of the uninucleate conidia into germ-tubes which have 
grown out of the liquid decoction, thus resulting in the formation at 
once of secondary conidia; jig. 7 shows the beginning of the forma- 
tion of a tertiary conidium. In figs. 8 and g the germ-tube has grown 
into a hypha in which all of the protoplasm appears to be in the 
end cells; in the latter, sixteen apparently empty cells separate the 
old conidial wall from the terminal protoplasm, which still remains 
uninucleate. 
‘Both BREFELD (’71, figs. 5, 29, 31) and THAXTER (’88, fig. 240) 
give figures showing a somewhat further. advance over what I have 
obtained in the cultivation of conidia, in that in their forms branches 
